In Southeast Asia’s startup market, the era of “raise fast, spend faster” has given way to a more disciplined question: how can companies keep growing without giving away too much of themselves?

That shift is creating room for financing products that sit between bank loans and venture capital. OSK Ventures International and Affin Hwang Investment Bank are now moving into that gap with the launch of Pothos Fund I, a dedicated venture debt fund aimed at high-growth companies across Southeast Asia.

Also Read: Venture debt in SEA: The non-dilutive capital that comes with hidden legal strings

The fund, managed through Pothos GP, has a three-year investment tenure and will provide debt-equity hybrid financing to companies that have moved beyond the earliest stage of startup life. Rather than backing ideas that are still being tested, Pothos Fund I will target revenue-generating businesses with proven models, stronger management teams and more predictable cash flows.

For founders, the appeal is straightforward. Venture debt can extend a company’s runway or fund expansion without forcing management teams to raise another equity round at an unfavourable valuation. For investors, the product offers exposure to private technology companies through a structure that includes contractual income, downside protection and selective equity participation.

The launch also gives sophisticated investors in Malaysia access to an asset class that has historically been more common among large institutional investors.

Why venture debt is becoming more relevant

Venture debt is not new, but it has become more visible as startups and investors reassess the cost of capital. In simple terms, it is a loan designed for venture-backed or high-growth companies that may not yet fit the credit models used by traditional banks. It is often paired with warrants or other equity-linked features, giving lenders some upside if the borrower performs well.

In Southeast Asia, the model has become more relevant for several reasons. The region’s digital economy has matured, producing more companies with recurring revenue, payment histories and expansion plans across multiple markets. At the same time, equity funding has become more selective after the global correction in tech valuations.

That combination has put pressure on founders to become more capital-efficient. Raising equity remains essential for many startups, especially those in capital-intensive sectors such as fintech, logistics, climatetech and artificial intelligence infrastructure. But for companies with clearer revenue visibility, debt can be a useful tool to finance working capital, product development, market expansion or acquisitions.

The timing is important. Southeast Asia’s startup ecosystem is no longer defined only by early-stage venture rounds. More companies now sit in the middle: too mature to be treated like seed-stage bets, but not yet large or profitable enough to borrow easily from commercial banks. It is this middle layer that venture debt funds are trying to serve.

What OSKVI and Affin Hwang bring to the table

The partnership combines OSKVI’s venture investing background with Affin Hwang’s capital markets and private markets structuring experience.

OSKVI, listed on Bursa Malaysia, has invested in, supported and exited more than 50 technology and enterprise companies across Southeast Asia over the past two decades. That history matters in venture debt, where lenders need to assess not only cash flow but also investor backing, founder quality, sector dynamics and the likelihood that a company can raise future capital if needed.

Also Read: Venture debt: How it stacks up against loans and equity

Affin Hwang brings a different set of capabilities, including fundraising, distribution, private markets structuring and access to institutional and sophisticated investors. Those strengths are useful at a time when wealth managers, family offices and other sophisticated investors in the region are looking for alternatives to public equities and traditional fixed income.

Pothos GP, the fund manager of Pothos Fund I, is a subsidiary of OSKVI, with strategic equity participation from Affin Hwang Investment Bank.

Amelia Ong, CEO of OSK Ventures International, framed the fund as part of a broader shift in how startups are financed.

“Having worked with entrepreneurs across Southeast Asia for many years, we have seen firsthand how access to the right capital at the right time can make all the difference,” she said. “As companies mature, their financing needs evolve, and venture debt provides a valuable option alongside traditional equity funding.”

A more crowded alternative capital market

Pothos Fund I enters a regional market where venture debt is still underdeveloped compared with the US or India, but no longer empty. In Southeast Asia, players such as InnoVen Capital, Genesis Alternative Ventures and AFG Partners have helped familiarise founders and investors with non-dilutive or less-dilutive growth capital.

Globally, the space includes specialist lenders such as Hercules Capital, as well as bank-linked providers such as HSBC Innovation Banking, which absorbed parts of Silicon Valley Bank’s operations outside the US after SVB’s collapse.

India offers a useful comparison for Southeast Asia. Over the past decade, venture debt firms such as Trifecta Capital and Stride Ventures have built sizeable businesses by lending to startups that had institutional equity backing and clearer paths to revenue. Southeast Asia has similar ingredients, but its market remains more fragmented, with startups operating across different regulations, currencies and customer behaviours.

That fragmentation can make lending harder. A startup expanding from Malaysia to Indonesia, Vietnam, or the Philippines faces different legal systems, payment rails and market risks. For venture debt funds, this means underwriting must go beyond a company’s balance sheet. It requires a view on the founders’ execution record, existing investor support, customer concentration, repayment capacity and the durability of demand.

The founder’s trade-off

Venture debt is often described as less dilutive, but it is not free money. Borrowers need to make repayments, and lenders typically include covenants or protections. If a company misses growth targets or burns cash faster than expected, debt can become a burden.

That is why funds such as Pothos Fund I are more likely to suit startups that already have revenue and a credible plan for cash generation, rather than early-stage companies still searching for product-market fit. Used well, venture debt can help a company avoid raising equity during a weak funding market. Used poorly, it can add pressure at precisely the moment a startup needs flexibility.

For Southeast Asian founders, the significance of Pothos Fund I lies less in the launch of a single fund and more in what it signals about the market’s direction. The region’s financing stack is becoming more layered. Equity capital remains important, but founders increasingly have more choices: revenue-based financing, venture debt, private credit, bank partnerships and strategic capital.

Also Read: Lighthouse Canton to offer access to venture debt to investors on Alta platform

That evolution is healthy. A mature startup ecosystem needs more than one type of money. It needs risk capital for bold ideas, growth capital for scaling businesses and credit products for companies that have earned the right to borrow.

Pothos Fund I is arriving at a moment when investors want more discipline and founders want more control. If it can find the right borrowers, it could help fill one of Southeast Asia’s persistent funding gaps: capital for companies that are growing up, but not yet ready to behave like traditional corporates.

The post Malaysia’s OSKVI and Affin Hwang move into venture debt with Pothos Fund I appeared first on e27.

In Southeast Asia’s startup ecosystem, founders are taught to focus on what can be measured: product velocity, fundraising, customer acquisition, growth metrics and operational scale.

These matter. But something deeper is quietly changing underneath them.

AI is collapsing the cost of competence. Products that once took years to build can now be replicated in months, sometimes weeks. Interfaces increasingly resemble one another. Messaging converges around the same language. Entire categories begin sounding interchangeable.

The result is not merely technological commoditisation. It is perceptual commoditisation. Even when companies are genuinely different, markets increasingly experience them as the same.

This is the real competitive crisis emerging in the AI economy. Most startups still believe they are competing at the layer of product. In reality, the battle has already shifted upstream, toward perception, interpretation and cognitive positioning.

Because in saturated markets, people do not evaluate deeply anymore. They filter aggressively.

Recognition replaces analysis. Familiarity replaces investigation. Cognitive shortcuts become survival mechanisms.

This is why traditional branding advice increasingly feels outdated. Brand is not a logo. It is not a visual identity. It is not social media aesthetics or clever taglines.

Those are surface artifacts. The real function of brand is environmental.

Brand shapes the interpretive conditions through which people decide what feels credible, relevant, trustworthy or important before conscious evaluation even begins. Before investors analyse metrics, before customers compare features, before talent evaluates offers, something has already shaped perception.

That perception influences whether people lean in or move on. Behavioural science has repeatedly shown that human decision-making is far less rational than most businesses assume. Daniel Kahneman’s work on cognitive shortcuts and heuristics demonstrated that people rely heavily on mental simplification when navigating uncertainty.

Also Read: Asian startups have an investor problem nobody is naming

AI amplifies this tendency because markets are now flooded with infinite information, infinite content and infinite comparison. The more options people encounter, the more they depend on interpretive shortcuts: trust, familiarity, clarity, narrative coherence, social proof, and perceived inevitability.

In other words, the future advantage is not merely visibility. It is interpretive control.

This is already visible in venture capital behaviour. Early-stage startups are routinely valued far beyond present-day financial performance because investors are not simply buying current capability. They are buying belief in future dominance.

That belief is shaped not only by technology or traction, but by whether a company feels culturally relevant, strategically inevitable and psychologically credible. This aligns with broader market data. Research from Ocean Tomo shows intangible assets now account for roughly 90 per cent of the market value of S&P 500 companies.

What markets increasingly value is not just operational capability. They value perceived defensibility.

Grab is a regional example of this dynamic. Much of its enterprise value comes not only from infrastructure or platform functionality, but from years of accumulated familiarity, behavioural trust and embedded relevance across Southeast Asia. That is not merely marketing. It is cognitive positioning at scale.

The same dynamic shapes pricing power. Companies with stronger perception resilience consistently command premiums even in highly competitive markets. Singapore Airlines continues to sustain premium positioning not solely because of operational performance, but because customers already associate the airline with reliability, confidence and quality before comparisons begin.

This is where most startup conversations about branding fail. They focus on expression instead of environment.

But in the AI economy, the companies that win will increasingly function less like products and more like worlds. The strongest companies build interpretive ecosystems that shape how people perceive reality around them.

Apple does not merely sell devices. It constructs a world around simplicity, taste and creative identity.

Nike does not merely sell shoes. It builds psychological associations around ambition, struggle and self-transformation.

The most powerful startups of the next decade will do something similar: they will shape meaning before evaluation starts.

This is where worldbuilding becomes commercially strategic rather than creatively abstract. Worldbuilding is the deliberate construction of signals, narratives, symbols, experiences and emotional triggers that create a coherent psychological environment around a company.

Also Read: Why so many startups are cutting down on the number of tools they use

Every interaction becomes part of the interpretive system: the founder’s language, the product behaviour, the onboarding experience, the interface, the hiring narrative, the investor story, the media presence, the customer community.

Together, these signals shape what people believe the company represents long before direct comparison takes place. In high-noise AI markets, this matters enormously. Because attention alone is becoming fragile.

AI-generated content has created an economy of infinite visibility but declining memorability. The startups that survive will not necessarily be the loudest. They will be the ones that reduce uncertainty fastest.

The ones that create cognitive ease. The ones that feel coherent under pressure. The ones that people instinctively understand and remember.

This is why emotional triggers matter more than many founders realise. Fear of irrelevance. Desire for belonging. Status signalling. Identity reinforcement. Risk reduction. Future aspiration.

The strongest companies understand that markets do not merely buy functionality. They buy emotional resolution.

Economist Robert Shiller described this dynamic as “narrative economics,” the idea that stories themselves shape economic behaviour and market outcomes. In the AI era, narrative becomes even more powerful because AI accelerates production faster than humans can process meaning. As sameness increases, interpretation becomes the new competitive frontier.

So what should startups actually do? The solution is not “better branding” in the traditional sense. It is strategic worldbuilding.

Founders need to stop asking: “How do we market our startup?” The more important question is: “What environment shapes how people perceive us before conscious evaluation begins?”

This changes how startups should think about growth entirely. Instead of treating brand as a late-stage marketing layer, startups should build interpretive infrastructure from the beginning: clarity of worldview, consistency of signals, narrative coherence, emotional resonance, behavioural trust, and psychological memorability.

Because in AI-saturated markets, the greatest threat is no longer invisibility. It is becoming cognitively interchangeable.

Over the next decade, many startups will fail not because their technology was weak, but because markets stopped perceiving meaningful differences between them. The companies that endure will not simply compete for attention. They will shape the environments through which attention becomes belief in the first place.

—

The post The new startup playbook: From product velocity to cognitive positioning appeared first on e27.

Ask a warehouse operator in Singapore what keeps them up at night, and forklifts come up before fires, floods, or fraud.

They should. Between 2022 and 2023, vehicular incidents were the leading cause of fatal workplace injuries in Singapore, and one in four of those deaths involved a forklift, as per the Ministry of Manpower (MOM).

MOM didn’t wait for the data to accumulate further. In November 2024, it introduced enhanced forklift refresher training requirements, on top of a 2015 circular on the safe use of storage racks, following fatal rack-collapse cases.

That’s the backdrop on warehouse safety in Singapore.

But here’s the forecast – within five years, every warehouse operating in Singapore will run on some form of AI-based safety monitoring. Not because it’s trendy. Because three things are converging at once, and none of them is slowing down.

The rules are no longer satisfied by good intentions

Singapore’s Workplace Safety and Health Act asks employers to take “reasonably practicable” steps to protect workers. For years, that meant a folder of risk assessments, a monthly walk-through, and a rack inspection schedule with daily visual checks, weekly compiled reports, and annual professional audits.

On paper, it works. In practice, a blocked emergency exit gets cleared for an audit and drifts back within days. PPE compliance holds in the morning shift and slips by the afternoon. A bent upright from a forklift impact goes unnoticed until the next scheduled inspection, weeks later.

Non-compliance isn’t a soft cost anymore. Fines under WSH regulations can run up to SG$500,000 for corporate entities, and severe violations trigger a Stop Work Order — a warehouse shutdown overnight, mid-fulfilment cycle. The Workplace Safety and Health Council has already named warehousing an accident hotspot, specifically around forklift use and loading operations. Regulators are asking for continuous, demonstrable compliance now, not a clean paper trail collected once a quarter.

That’s a standard periodic inspection that was never built to meet — and it’s precisely the standard AI-based monitoring is built to meet instead, because it doesn’t inspect on a schedule. It watches continuously, which is the only way “reasonably practicable” starts to mean something real rather than something documented after the fact.

Also Read: Why your data warehouse is just a very expensive attic

The floor is shrinking while the volume grows

Singapore’s freight and logistics market is worth roughly US$26 billion this year and is on track to hit over US$35 billion by 2031, growing at more than 6 per cent annually, with warehousing itself among the fastest-growing segments, propelled by e-commerce and just-in-time stocking demand.

That growth is landing on a footprint that isn’t expanding at the same rate. Land is scarce and expensive. Warehouses are going vertical, consolidating operations, and running leaner headcounts than the volume suggests they need. A supervisor who once covered one aisle now effectively covers three.

More product moving through less space, watched by fewer people, is a formula periodic manual checks were never designed to handle, and it’s exactly the gap AI is being built to close. A camera system that already exists on-site doesn’t need more headcount to watch more aisles; it just needs to be given the job.

The technology stopped being the limitation

The biggest change, if we consider the last five years in warehouse safety technology, is not that the cameras have become better. It is that the purpose of the camera has evolved.

For most warehouses, CCTV has historically been a forensic tool. Footage becomes valuable after something has happened, like an injury, a collision, damaged stock or a disputed near miss. Someone identifies the approximate time, retrieves the recording and reconstructs the event.

Computer vision-based monitoring changes that sequence.

Instead of waiting for a supervisor to review footage, AI models can analyse visual conditions as operations unfold and identify predefined risk patterns. That distinction matters because many warehouse incidents develop over seconds rather than hours, leaving very little time for conventional supervision to intervene.

The technical barriers to doing this at operational scale have also fallen. The AI warehouse monitoring systems can increasingly work with existing surveillance infrastructure like CCTVs on site, while edge computing allows safety-critical processing to happen close to where footage is generated rather than depending entirely on cloud connectivity.

But detection itself may prove to be only the first stage.

Warehouses generate thousands of visual observations across shifts, aisles and loading areas. Analysed over time, those observations can reveal something more valuable than individual violations, for example, the patterns of exposure.

Also Read: Boardrooms to warehouses: How SEA leaders can build cyber resiliency from top-down

This is where newer technological developments like vision-language models (VLMs) and agentic AI systems could push warehouse safety further. Rather than simply classifying an event, these systems are beginning to interpret sequences of activity, retrieve relevant evidence and help safety teams identify recurring conditions across larger volumes of operational data.

That changes the role of collected footage on the floor again. It translates from evidence of what happened to detection of what is happening, and eventually to intelligence about what is likely to keep happening unless the underlying condition changes.

Warehouses generate an enormous volume of operational data every second, but most of it has traditionally gone unused because it couldn’t be analysed in real time. AI changes that by transforming visual information into measurable safety intelligence, allowing organisations to intervene before isolated events develop into systemic risks.

Five years is the generous estimate

None of this replaces a supervisor’s judgment or a good toolbox talk. What it removes is the lag between a hazard forming and someone catching it — where most warehouse incidents live.

Put the three forces together — regulators demanding continuous proof, a market outgrowing its floor space and headcount, and AI infrastructure finally cheap and local enough to run on cameras a warehouse already owns — and five years starts to look conservative, not ambitious.

The operators moving now aren’t betting on a trend. They’re the ones who read the regulatory notices, looked at the growth numbers, and did the math first.

—

The post Why every warehouse in Singapore will run on AI safety monitoring within five years appeared first on e27.

The environmental debate around AI is often placed in the sustainability section of the company, where it becomes a reporting matter, a disclosure matter, or a reputational matter. By the time it gets there, most of the important decisions have already been made.

The environmental impact of AI is not shaped mainly by the annual report. It is shaped by product choices made much earlier. Which model was selected. How often it is called. Whether the system defaults to generation when retrieval would do. Whether latency targets force expensive compute. Whether every user action triggers inference or only the ones that matter. Whether the team built a feature that solves a real problem or simply adds a layer of fashionable intelligence to something that was already working.

Every model choice is also a resource choice

A strange habit has developed in many AI teams. Model choice is discussed as though it were mainly a question of quality, capability, or technical ambition. Bigger model or smaller model. Faster model or smarter model. In practice, that decision also carries implications for cost, latency, infrastructure strain, and carbon impact.

If a team chooses a heavier model for a use case that only needs a narrower and cheaper one, that is not only an architecture decision. It is a product judgement. The team has decided that the extra compute is justified by the user value. In many cases, nobody says it that directly, which is exactly why the ethics stay fuzzy.

Latency pressure can become an ethical problem, not just a product one

There is another layer that companies do not examine closely enough. The modern product instinct is to push for lower latency at almost any cost. Faster feels better. Faster looks more advanced. Faster improves adoption and makes the feature feel magical.

But in AI systems, faster can also mean more expensive infrastructure choices, more aggressive provisioning, less efficient batching, and more resource-hungry serving patterns. A company may think it is making a user experience decision when it insists on near-instant generation everywhere. In reality, it may be making a hidden decision about energy intensity and carbon burden for a very marginal gain in perceived user delight.

Also Read: How to get beyond the chatbot and boost your AI productivity

A serious team should be able to ask a harder question. Does this use case truly require this speed, or are we burning more compute to remove a few seconds of waiting that users would have accepted quite happily? That is not anti-innovation. It is disciplined judgement.

Carbon is often the result of weak product discipline upstream

Many firms speak about AI emissions as though they are the unavoidable byproduct of progress. That framing lets product teams off too easily. A large share of the environmental cost in AI is not simply the price of doing business. It is the price of design choices that were never challenged properly.

Consider how much waste enters the system through product habits that are treated as normal. Features that call a model too often. Workflows that trigger repeated generation because the first output is not grounded well enough. Interfaces that encourage users to regenerate endlessly because nobody designed for confidence or finality. Architectures that use large models for routine classification or extraction tasks. Orchestration layers that look sophisticated but create multiple expensive calls where one would have been enough.

None of this is abstract. It is the operational reality of many AI products.

When viewed that way, environmental ethics starts looking less like a sustainability speech and more like a test of product seriousness. Teams that cannot control unnecessary inference, retries, and overbuilt flows are not only weak on cost discipline. They are weak on environmental discipline too.

The most responsible AI products will not always be the most technically flamboyant

There is still too much status attached to using the most powerful model available. In some companies, restraint is interpreted as compromise. Smaller models look less ambitious. Simpler architectures look less impressive. Retrieval-first systems can sound less glamorous than generative ones. But the product leader with mature judgement will increasingly ask a more grounded question.

What level of intelligence is actually required for this task?

That question matters because many business problems do not need the full weight of frontier capability on every interaction. Some tasks need reasoning depth. Some need consistency. Some need structured extraction. Some need speed. Some need a safe and bounded answer. Treating all of them as invitations for maximum model power is not thoughtful design. It is often a failure to match compute intensity to user value.

Cost, carbon and user value should be discussed together, not separately

One reason this issue remains weakly governed is that organisations split the conversation into silos. Product talks about user benefits. Engineering talks about performance. Finance talks about cost. Sustainability talks about carbon. By the time those views meet, the feature is usually already live, and the room is arguing over trade-offs that were baked in earlier.

This is the wrong sequence.

Also Read: Product management as method acting: Becoming your user

A stronger company would ask these questions together from the beginning. What value is this feature creating? What is the latency expectation that truly matters? What is the marginal gain from using a more expensive model? What does that do to operating cost at scale? What does it imply for resource consumption? Is there a lighter path to the same user outcome?

Environmental ethics should shape the product brief, not the corporate statement

The most distinctive shift companies need to make is procedural. Environmental responsibility in AI should be built into the product brief itself.

A team should be able to explain why this model class is appropriate for this job. Why this latency level is worth the infrastructure burden. Why is this frequency of inference necessary? Why this workflow cannot be narrowed? Why does this user need to justify this operational intensity? If the team cannot answer those questions clearly, then the sustainability language that appears later is unlikely to mean very much.

This is what makes the issue strategic rather than symbolic. Product leaders decide what gets built, how much complexity gets added, what kind of performance is pursued, and where efficiency is allowed to shape the experience. Those decisions are environmental decisions whether they are written that way or not.

A company that leaves this entirely to sustainability reporting is effectively saying it wants to measure the consequence without governing the cause.

The next generation of strong AI products will look more selective

There is a common assumption that the future belongs to products that apply AI more broadly and more aggressively. In practice, the stronger products may be the ones that apply it more selectively and more intelligently.

They will know where generation is truly useful and where deterministic systems are better. They will know where latency matters and where patience is acceptable. They will know when to reserve heavy models for exceptional cases rather than routine flow. They will treat inference as something to allocate deliberately, not something to spray across the interface because it feels innovative.

That kind of selectivity will produce better economics, better operational control, and a cleaner environmental posture. More importantly, it will reflect a better philosophy of product building.

—

The post The environmental ethics of AI should be a product decision, not a sustainability footnote appeared first on e27.

Strengths. Weaknesses. Opportunities. Threats.

Someone fills four boxes with familiar phrases, takes a photo, and never looks at it again.

That is not a strategy framework. It is office wallpaper.

Used at the right time, however, frameworks such as 5W1H, SWOT, and PESTLE can help a person avoid one of the most expensive mistakes in innovation: building the wrong thing with great enthusiasm.

This matters more in the age of AI.

AI can produce 50 product ideas before lunch. It can turn a rough thought into a neat business plan, a product description, and a list of potential customers. It can make an early idea look much more finished than it really is.

The problem is not a shortage of possibilities. The problem is deciding which possibility deserves your time.

That is what frameworks are for.

Start with the problem, not the solution

The first framework is also the simplest: 5W1H.

Who has the problem? What exactly happens? When does it happen? Where does it happen? Why is it costly or frustrating? How do people cope with it now?

These questions sound obvious. They are not.

Many weak ideas begin with a solution looking for a problem. Someone wants to use AI, build an app, add a sensor, or invent a feature. Then they go searching for a reason to justify it.

5W1H flips the order. It forces the inventor or business owner to describe the real situation first.

Consider a restaurant owner who says, “I need an AI tool for stock management.” That is a solution. The better question is: what is actually going wrong? Is food being wasted because demand changes? Is staff input unreliable? Are suppliers late? Is the problem fresh ingredients, storage, purchasing, or forecasting?

Also Read: Can AI really improve collaboration and productivity

The answer changes what should be built.

A good problem statement does not make an idea less creative. It gives creativity a target.

Use SWOT before the money is spent

SWOT is most useful after you have a possible solution but before you have committed too much time or money.

A strength is not just something you are proud of. It is an advantage you can use. A weakness is not an admission of failure. It is a constraint that may shape the first version. An opportunity is not a vague trend. It is a change you can act on. A threat is not a reason to give up. It is a risk you need to design around.

Imagine an SME that has developed a better way to inspect a component before it leaves the factory.

Its strength may be deep knowledge of the production line. Its weakness may be limited software skills. Its opportunity may be rising demand for traceability. The threat may be that a large global supplier can quickly copy a visible feature.

That last point is important. A SWOT analysis can lead directly to an IP question. If the innovation is easy to see and valuable, should the company explore patent protection? If the value sits inside a hard-to-observe process, should it be kept confidential as a trade secret?

The framework does not answer the question. It makes sure you ask it while there is still time to act.

PESTLE helps you see the weather

PESTLE looks outside the business: political, economic, social, technological, legal, and environmental forces.

It is easy to dismiss as another consultant’s acronym. That would be a mistake.

A good idea can fail because it arrives at the wrong time, in the wrong market, or under the wrong rules. A PESTLE scan helps you notice the weather before you set sail.

A product may be timely because regulations are changing. A new solution may struggle because customers are cutting costs. Climate pressure may create demand for less waste. A shift in trade rules may make local alternatives more valuable. An aging population may create a need for a different kind of service.

These are not background details. They shape whether an invention has a market.

For Southeast Asian businesses, this matters because the region contains many different markets. A solution that works in Singapore may need a different price, partner, or compliance path in Indonesia, Vietnam, Thailand, or the Philippines.

A framework is a set of better questions

The point is not to complete three templates and declare yourself innovative.

A framework is useful when it turns a fuzzy idea into a sharper question.

5W1H asks whether you understand the problem. SWOT analysis asks whether your solution aligns with your real strengths and risks. PESTLE asks whether the external environment is helping or hindering your timing.

Also Read: From copilots to colleagues: How agentic AI is redefining enterprise productivity

Together, they can turn an exciting thought into a practical experiment.

What do we need to test first? Who should we talk to? What would prove we are wrong? What part of the idea creates the value? Could a competitor easily copy it? What should stay secret? What might be worth protecting?

This is where frameworks become more than management language. They become a bridge between a bright idea and a decision.

AI needs a good brief too

AI is most useful when it is given good context. A vague prompt produces a vague answer, even when it sounds confident.

If you use 5W1H to define the problem, SWOT to understand the business, and PESTLE to see the market conditions, you can give AI a much better brief. Then it can help generate options, compare approaches, identify questions, and organise research.

It can make the thinking faster.

It cannot make the thinking optional.

The world does not need more beautifully presented ideas that fail the moment they meet reality. It needs more people who can turn a real problem into a clear, tested, and defensible solution.

That is not boring.

That is how innovation gets built.

If you have an idea that needs sharper questions before it needs a big budget, start exploring it for free at IPGuru.ai.

—

Editor’s note: e27 aims to foster thought leadership by publishing views from the community. You can also share your perspective by submitting an article, video, podcast, or infographic.

The views expressed in this article are those of the author and do not necessarily reflect the official policy or position of e27.

Join us on WhatsApp, Instagram, Facebook, X, and LinkedIn to stay connected.

The post SWOT is not boring; you are just using it too late appeared first on e27.